Steve Appleton, Micron Chairman, CEO, and President, testified from about 10
a.m. until about 6:20 p.m. with an hour lunch break and ten minute breaks
every two hours or so. The session began with ALJ McGuire distributing an
order to both sides, telling them to read it, and let him know when they had.
My observations during the session are consistent with the Hager conclusion
that this was the order regarding how testimony regarding collusion would be
treated, specifically that parties would be allowed to conduct limited
questioning with regard to the DOJ investigation.

CC Royall and the witness spent almost five hours to establish what they
wanted on the record: (1) Mr. Appleton's meteoric rise through the ranks of
Micron Corporation after graduation with a degree in business management from
Boise State College, which he attended on a tennis scholarship, from
production operator in 1983 to, eleven promotions later, president in 1991,
and subsequently CEO and chairman. Along the way, the point was made that he
is not an engineer. (2) Micron's evolution over only a slightly longer time
from a consulting company with only five employees to one of the world's
largest DRAM manufacturer's with 17,000 employees. Rambus Counsel Greg Stone
objected to the dilatory exchange but was overruled as Mr. Royall claimed
that the background is necessary for consideration of remedy which has not
been addressed. (3) All of Micron's products conform to Jedec standards, that
these are critical to customer and manufacturer. When all can design to a
single standard it will be cost-effective and the customer will know he has a
variety of suppliers. It did not require an objection here for Judge McGuire
to tell Mr. Appleton to refrain from making speeches. (4) The DRAM industry
is down to five or six major players, one in each of the major areas of the
world, Micron being the only US-based manufacturer, although others have
manufacturing facilities in the US. That the DRAM market, a commodity
business because of open standards that permit entry of manufacturers, is the
most volatile of all markets in both supply and demand, and that Micron has
managed to remain competitive by focusing on cost, innovation, and moving
forward, reducing cost/bit 25-30% annually, and this year aiming for a 50%
reduction. (5) Micron has had a string of losses, four of the last five
years, and reported a one-billion dollar loss this year. Mr. Stone objected
because Micron would not produce cost information to Rambus counsel making
this an inappropriate line of questioning. He cautioned that Rambus counsel
would revisit this and recall Mr. Appleton even though his attorney had told
Rambus counsel that he will not return "because it is too expensive to
travel". Mr. Royall said that financials were nothing more than was contained
in annual reports.

(6) Micron has positioned itself for the future based on four core
principles: (i) innovation (ii) cost efficiency (iii) design to open
standards to increase market penetration (iv) customer focus. Micron has done
a "fabulous job" in capturing its own IP. R&D has increased 7-fold during
difficult times becoming the #3 company in the US for patents received, with
a current patent portfolio of 10,000 and 1,600 granted in 2002. Capturing its
own IP has allowed Micron to reduce royalty payments from 10% of its revenue
to significantly less than one percent and has put it in a good position in
cross-licensing negotiations. The company has never refused to pay royalties,
however. The company has controlled costs by addressing size (believe he was
talking about wafer size and process technology) even as the production cycle
has shortened from two to two-and-a-half years in 1994 to the current twelve
to eighteen months, extending the useful life of equipment through process
innovations. (7) Micron prefers to exist in an evolutionary rather than a
revolutionary environment because introduction of something very different
disrupts manufacturing. Evolution allows longer equipment lifetimes. This
works to the customers' advantage, too, because it makes product and supply
of product more reliable. (8) Micron reduced its royalty expenses from 10% of
revenues to significantly less than one percent by negotiations, though the
company has never refused to pay royalties. (9) Open standards mean that a
large body of industry participants use the same platform to move forward
with a minimum of royalties. Manufacturers benefit from open standards
because greater market penetration results. (10) Jedec is an open standards
organization. It provides for customers and competitors to meet and discuss a
shared standard. There is an expectation that no one afterwards will block
you. Judge McGuire: That's enough! (not loudly, in fact, I'm not sure I even
heard this entirely correctly, but it was said in a peremptory manner.) (11)
Customer alignment, MM's and customers sharing their roadmaps, timelines of
products and quantities over several years, is allowed. MM's want to build
what customers will use. Customer roadmaps are ever changing because of
unpredictability of product adoption and technical difficulties in producing
product. At this point, a multipage document, Winning DRAM Implementation
Game" Application Space", was projected. (I suggest that anyone who attends
future presentations bring opera glasses because it is otherwise difficult to
get a good view of projected material.) Mr. Appleton said it is the kind of
thing used with customers. It describes the nature of devices and timing of
products and is important for internal co-ordination, too. Confidential
proprietary information is often shared with customers in these
presentations. This document conveyed that Micron planned to produce all
different types of DRAM, including SDRAM and RDRAM. Micron solicits feedback
for future incorporation but also seeks to promote its own solutions as it
did with Burst EDO over SLDRAM in the early to mid-nineties because it
believed that it transferred data faster and at lower cost. However, it was
not standardized, so it was only sampled and never manufactured. In
response to Mr. Royall's question of "Why not?", Mr. Appleton said,
"Because the customer decides and it didn't address the customers' longer
term desires."


R: Were you working on other types of DRAM?
A: Multiple types. It's never an option to work on just one. We simultaneously
   developed other DRAM's along with Burst EDO.
R: Why multiple types?
A: The customer decides. We need to supply what the customer wants.
R: Are there examples of other memories that were not produced?
A: SyncLink and Rambus.
R: Why no production on SyncLink?
A: Customer decided against it.
R: And Rambus?
A: Depends on timing, what the customer wanted. Ultimately because customer didn't want to adopt it.
A: (to the judge) I didn't want to ramble. I just wanted to give the answer.
A: We promoted SyncLink and RDRAM.
R: Ever express concerns or reservations to customers?
A: Promoting Burst EDO w didn't think SDRAM was most cost effective. Burst EDO
   would have been an evolutionary step. Despite our reservations, the market
   wanted SDRAM. We made it.
R: Do customers solicit your reservations?
A: Yes.
R: Ever withhold reservations?
A: No. We think trust is a critical part of our relationship.
R: (to the judge) This might be a convenient breaking point.
J: How much more of your examination?
R: About one-and-one-half hours.

Court recessed for lunch.

Afternoon continuation of CC Royall's direct examination of witness Appleton

R: Micron is a Jedec member?
A: At least since I was president. Micron is supportive of and benefits from
   Jedec. Customers participate, too. . . . I've never attended a Jedec meeting
   or viewed minutes. My knowledge derives from Micron's representatives.
R: As CEO, is it important for you to be involved?
A: Yes, because activities affect the standards.
R: Your understanding of Jedec purpose?
A: Develop open standards for companies and customers to have access for product development.
R: How does Jedec approach open standards?
A: Share knowledge with participants as standard gets developed so knowledge
	 couldn't be used later against the standard.
R: Your views on your company's participation on intellectual property?
A; If we have IP that affects the standard, we should disclose. We wouldn't
   try to enforce if it did.
R: Are you aware of Micron enforcing a patent which should have but wasn't disclosed?
A: No.
R: Does others' behavior with regard to patents concern you?
A: Yes. It could affect the process and whether Micron participated, because
   it destroys the ability to set standards.
R: On the topic of Rambus, you first heard of them when?
A: About 1996 timeframe. Some technical folks were aware of the interface
   technology RDRAM for DRAM. It addressed the speed at which data can be moved
   in and out of memory.
R: Was Rambus making any cost-related claims?
A: Rambus claimed not much of a premium compared to other technologies.
R: And Micron's response?
A: Skepticism at their performance claims. They hadn't manufactured. We had a
   lot of manufacturing experience. Also, we were skeptical of their cost claims.
R: When were you personally contacted by Rambus?
A: I believe later in 1996.
R: Exhibit RX828
A: A letter from Geoff Tate, Dec 1996. I think this was the first written communication.
R: What was Rambus seeking?
A: Micron to take a license to RDRAM.
R: The sentence beginning "Our list price. . . ." referring to direct RDRAM--
A: Their technology. We thought the royalty was high (2.5%) and it was
   high-cost generally (implementation package $9.5 mil). Interface is such a
   small part of the eventual product. We thought the rate was very high.
R: You had paid DRAM royalties before. What was the difference here?
A: We had paid broad cross-licenses before. For example, Texas Instruments,
   they invented the semiconductor. We never paid licensing fees for memory
   interface technology.
R: High-speed development devices?
A: The letter itself says high-bandwidth DRAMS.
R: Was Micron considering other such technologies?
A: Increasing SDRAM speed, SyncLink.
R: Was Rambus asking anything with regard to other technologies?
A: Yes, stop working on them. "Put all your wood behind one arrow, RDRAM."
   (Don't think Mr. Appleton actually spoke these words. The phrase appears in
   the Rambus letter to Micron and is apparently Geoff Tate's wording.)
R: Your response?
A: We never take just one option. We can't limit ourselves. And we thought
   there were better alternatives.
R: Did Intel's thoughts affect yours?
A; Yes. Once Intel developed RDRAM, it increased the likelihood of customer adoption.
R: A license was discussed with Rambus?
A: Yes.
R: You participated?
A: Yes, in 1997, with Bob Donnelly, Vice-President, and someone from legal.
R: Who from Rambus?
A: Geoff Tate and a couple of others.
R: Did you take a license?
A: Yes.
R: Was there more than one meeting?
A: Yes. I don't know how many. I didn't participate in all.
R: A license was signed?
A: In late spring 1997.
R: Why?
A: With Intel's endorsement there would a customer base.
R: Did you stop working on other options?
A: No.
R: Did the license require that?
A: No.
R: You understood Rambus claimed to have patents to RDRAM technology?
A: Yes.
R: Do you recall patents or applications being provided to you?
A: No.
R: Did Rambus say its technology extended beyond RDRAM?
A: Never.
Judge: I don't understand. You licensed without seeing patents?
A: Yes, but we understood we would have access to patents.
R: Did Rambus ever indicate patents applied to anything other than RDRAM?
A: Never.
R: If it was the case, would you expect Rambus to bring it up?
Stone: Objection. Speculation.
Judge: Sustained.
R: Did Rambus have reason to know Micron was working on SDRAM?
A: Yes. It was referred to in the letter that we stop working on
   high-bandwidth DRAM after EDO SDRAM.
R: If Rambus possessed or believed it possessed patents over SDRAM, would it
   have been important to you?
A: Yes.
R: Did Rambus express belief it could have patents over SyncLink?
A: No. Knowledge of this would be important.
R: You put things in place for RDRAM?
A: All the things you do, a large design team, the largest of any product we
   had. When we signed the licensing agreement the final specifications were not
   complete but we put things in place--cash bonuses to the team for a host of
   RDRAM mileposts--this is unusual in Micron's approach--roadmaps with RDRAM,
   advertising for it.
R: Did you issue press releases?
Exhibit: June 1999 press release
A: This one announced 128 and 144 Mb RDRAM samples were shipped to Intel.
R: Was this the only press release?
A No, there were others.
They review other aspects of this press release.
R: You considered efforts successful to this point?
A: Yes.
R: You advertised about RDRAM?
A: Yes, print advertising.
R: What was said?
A: High-speed, high-bandwidth.
R: What did you communicate to customers?
A: That it was higher cost.
R: And performance?
A: Yes, we presented all the alternatives to performance. We believed SDRAM
   could be extended.
R: You did not take RDRAM into volume production?
Judge: Clarify volume.
Discussion ensues on sampling volume and production volume.
A: We never commercialized the product.
R: Why?
A: An accumulation of many events: delays in development, changes in design
   even after we expected it to be stable, customers who had forecast it but with
   difficulties and delays changed forecasts. Eventually, there was such a low
   level we decided not to commercialize.
R: Did others produce?
A: Samsung.
R: Did you reconsider?
A: Yes, it was a tough decision. One of our primary competitors produced it,
   but customer preference changed and we decided against it.
R: Any particular customer?
A: Dell was one of our largest. In our first discussions with Dell, they were
   very aggressive in demand. Then came a time when Dell said it was getting all
   the RDRAM they needed.
R: About the late nineties' Micron agreement with Intel:
A; Intel invested in Micron, a primary investment of 500 million to purchase
   advance process technology equipment. We would match them dollar for dollar.
R: Were you involved?
A: I was. I initiated them with a call to Dr. Barrett himself.
R: Why?
A: We were in the process of acquiring Texas Instruments' memory operations
   then and figured Intel would be interested in supporting continuing DRAM
   production.
R: Why Intel and not the bank?
A: Intel was interested in continuing prod
R: Were conditions linked to prerequisites?
A: Yes, Intel RDRAM-compatible chipsets and timing of development.
R: Why was the agreement structured with these conditions?
A: The customer decides. We had them put conditions in to avoid them making something that wouldn't sell.
R: Were conditions that trigger Micron production ever met?
A: No.
R: Is the agreement still in effect?
A: No.
R: It ceased when?
A: (I missed the reply.)
R: Did Micron meet its commitments?
A: Yes.
R: Did you hear from Intel that you didn't?
A: No.
R: Micron supported SLDRAM?
A: Yes, another approach the market would adopt.
Exhibit RX801, a letter from Tabrizi, Hynix marketing executive, to Steve Appleton.
R: You learned Rambus asserted patents against SDRAM and DDR?
A: Yes, from the Hitachi suit in 2000. We were surprised and disappointed
   because both SDRAM and DDR were the result of an open-standard Jedec process
   and now an IP company was asserting a right to it. . . . Rambus never
   mentioned in RDRAM licensing discussions that its IP applied to SDRAM or other
   technology being developed. (He became very intense in saying this.) . . . We
   were concerned because we had made an investment. It's not an easy change for
   us to make. Our process and customer base has designed for it. . . . Rambus
   claimed in the press it was asking exorbitant royalty rates for SDRAM and DDR
   and might not license. . . .We were contacted--a letter from Tate--the latter
   half of 2000 describing terms of how we should take a license.
Stone: Objection. The letter is the best evidence of its contents.
Judge: Sustained.
R: It contained?
A: Royalty rates.
R. Did you meet with them?
A: No?
R: The reason?
A: Internal discussions. We were disappointed and pretty upset for the simple
   reason that they should have told us during earlier discussions. They
   participated in discussions of development of standards (becomes intense).
Judge: You are becoming argumentative.
R: Rambus was seeking royalties?
A: One percent on product currently being made (SDRAM), 3.5% on future 
   production (DDR). (Witness was indignant in recounting this.)
R: Your opinion?
A: We thought they were exorbitant.
R: The cost to Micron of these royalties?
A: Hundreds of millions of dollars.
Exhibited an earlier slide
R: And that would affect you how?
A: That would be 25-50% of our R&D expenditures.
R: Did Micron agree to the license?
A: No.
R: Why not?
A: Somebody has to stand up and do the right thing. The royalties were too
   high. The process was already in place.
R: And this put your company at risk?
A: Yes. Rambus might not license on its patents and it would affect our
   ability to produce.
Exhibit: and email exchange between Appleton and Tate
(Don't know just what's what here, A's testimony or email contents.)
not taking license
reference to Rambus past and recent behavior. (Rambus' suits against Hitachi and another.)
A: We thought it was bad faith. The best course was to let the courts decide.
   It was inevitable.
R: Micron instituted litigation against Rambus?
A: Not then, later, against a declaratory judgment against patents. We
   asserted violation of antitrust, fraud, and bad faith.
R: Was this the only suit?
A: No, there were suits in France, Germany, Italy, and the UK.
R: The nature of the suits?
A: Asserting for patents.
R: Are foreign suits a risk?
A: Absolutely. Rambus tried to get a preliminary judgment to shut us down in Italy.
R: How would that affect you?
A: It would become an unproductive investment. There would be unemployment. We
   would be unable to service customers.
R: And Germany?
A: Same story (my words, he was more wordy.)
R: The last topic: after you learned of Rambus asserting patent rights against
   SDRAM and DDR, did you phase out production?
A: No, you don't want to move sideways after investing; you want to move forward.
R: Because of the customer base?
A: The customer has developed products based on standards. We can't change
   until they are done with it.
R: Design qualification?
A: The design time is one to two years for changing, then sampling for six to
   nine months, then get it qualified three to five months, then get customers to
   adopt.
R: So you can't move sideways?
A: We must offer an advantage to the customer.

Conclusion of CC direct examination of Mr. Appleton